Watts Up With That? 
Oct 27, 2010
From Warren Meyer at climate-skeptic.com  From the Thin Green Line, a reliable source for any absurd science that supports environmental alarmism:
Sending and receiving email makes up  a full percent of a relatively green person’s annual carbon emissions, the equivalent of driving 200 miles.
Dealing with spam, however, accounts for more than a fifth of the average account holder’s electricity use. Spam makes up a shocking 80 percent of all emails sent, but most people get rid of them as fast as you can say “delete.”
So how does email stack up to snail mail? The per-message carbon cost of email is just 1/60th of the old-fashioned letter’s. But think about it — you probably send at least 60 times as many emails a year than you ever did letters.
One way to go greener then is to avoid sending a bunch of short emails and instead build a longer message before you send it.
This is simply hilarious, and reminds me of the things the engineers would fool the pointy-haired boss with in Dilbert. Here was my response:
This is exactly the kind of garbage analysis that is making the environmental movement a laughing stock.
In computing the carbon footprint of email, the vast majority of the energy in the study was taking the amount of energy used by a PC during email use (ie checking, deleting, sending, organizing) and dividing it by the number of emails sent or processed. The number of emails is virtually irrelevant — it is the time spent on the computer that matters. So futzing around trying to craft one longer email from many shorter emails does nothing, and probably consumers more energy if it takes longer to write than the five short emails.
This is exactly the kind of peril that results from a) reacting to the press release of a study without understanding its methodology (or the underlying science) and b) focusing improvement efforts on the wrong metrics.
The way to save power is to use your computer less, and to shut it down when not in use rather than leaving it on standby.
If one wants to argue that the energy is from actually firing the bits over the web, this is absurd. Even if this had a measurable energy impact, given the very few bytes in an email, reducing your web surfing by one page a day would keep more bytes from moving than completely giving up email.
By the way, the suggestion for an email charge in the linked article is one I have made for years, though the amount is too high. A charge of even 1/100 cent per email would cost each of us about a penny per day but would cost a 10 million mail spammer $1000, probably higher than his or her expected yield from the spam.